Assessment of respiratory flow cycle morphology in patients with chronic heart failure

详细信息    查看全文

• 作者：Ainara Garde ; Leif Sörnmo ; Pablo Laguna…
• 关键词：Chronic heart failure ; Respiratory pattern ; Periodic and non ; periodic breathing ; Ensemble average
• 刊名：Medical & Biological Engineering & Computing
• 出版年：2017
• 出版时间：February 2017
• 年：2017
• 卷：55
• 期：2
• 页码：245-255
• 全文大小：
• 刊物类别：Engineering
• 刊物主题：Human Physiology; Biomedical Engineering; Imaging / Radiology; Computer Applications;
• 出版者：Springer Berlin Heidelberg
• ISSN：1741-0444
• 卷排序：55

文摘

Breathing pattern as periodic breathing (PB) in chronic heart failure (CHF) is associated with poor prognosis and high mortality risk. This work investigates the significance of a number of time domain parameters for characterizing respiratory flow cycle morphology in patients with CHF. Thus, our primary goal is to detect PB pattern and identify patients at higher risk. In addition, differences in respiratory flow cycle morphology between CHF patients (with and without PB) and healthy subjects are studied. Differences between these parameters are assessed by investigating the following three classification issues: CHF patients with PB versus with non-periodic breathing (nPB), CHF patients (both PB and nPB) versus healthy subjects, and nPB patients versus healthy subjects. Twenty-six CHF patients (8/18 with PB/nPB) and 35 healthy subjects are studied. The results show that the maximal expiratory flow interval is shorter and with lower dispersion in CHF patients than in healthy subjects. The flow slopes are much steeper in CHF patients, especially for PB. Both inspiration and expiration durations are reduced in CHF patients, mostly for PB. Using the classification and regression tree technique, the most discriminant parameters are selected. For signals shorter than 1 min, the time domain parameters produce better results than the spectral parameters, with accuracies for each classification of 82/78, 89/85, and 91/89 %, respectively. It is concluded that morphologic analysis in the time domain is useful, especially when short signals are analyzed.